A Precise Program Phase Identification Method Based on Frequency Domain Analysis

09/10/2021
by   Hsuan-Yi Lin, et al.
0

In this paper, we present a systematic approach that transforms the program execution trace into the frequency domain and precisely identifies program phases. The analyzed results can be embedded into program code to mark the starting point and execution characteristics, such as CPI (Cycles per Instruction), of each phase. The so generated information can be applied to runtime program phase prediction. With the precise program phase information, more intelligent software and system optimization techniques can be further explored and developed.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
04/25/2023

TraceFixer: Execution Trace-Driven Program Repair

When debugging unintended program behavior, developers can often identif...
research
03/04/2016

Performance Localisation

Performance becomes an issue particularly when execution cost hinders th...
research
12/13/2021

Public Release and Validation of SPEC CPU2017 PinPoints

Phase-based statistical sampling methods such as SimPoints have proven t...
research
11/08/2022

Final Report on MITRE Evaluations for the DARPA Big Mechanism Program

This report presents the evaluation approach developed for the DARPA Big...
research
07/24/2021

SODA: A Semantics-Aware Optimization Framework for Data-Intensive Applications Using Hybrid Program Analysis

In the era of data explosion, a growing number of data-intensive computi...
research
07/16/2019

A Survey of Phase Classification Techniques for Characterizing Variable Application Behavior

Adaptable computing is an increasingly important paradigm that specializ...
research
01/23/2021

Recovery and Analysis of Architecture Descriptions using Centrality Measures

The necessity of an explicit architecture description has been continuou...

Please sign up or login with your details

Forgot password? Click here to reset